In the present years commercialization efforts for the synthesis of complex carbohydrates including secreted oligosaccharides have increased significantly due to their roles in numerous biological processes occurring in living organisms. Secreted oligosaccharides such as human milk oligosaccharides are becoming important commercial targets for nutrition and therapeutic industries. However, the syntheses and purification of these oligosaccharides and their intermediates remained a challenging task for science. One of the most important human milk oligosaccharides is 2′-O-fucosyllactose (α-L-fucopyranosyl-(1→2)-β-D-galactopyranosyl-(1→4)-D-glucose, “2′-FL”) found in the highest concentration in mother's milk.

Several biological roles of 2′-O-fucosyllactose have been suggested including but not limited to its prebiotic, antibacterial, antiviral, immune system enhancing, brain development enhancing, etc. effects making it an attractive target for large scale production/isolation/purification for nutritional and therapeutic industries. 2′-O-fucosyllactose has been synthesised by both chemical and enzymatic methodologies but commercially attractive production processes have not been developed due to lack of efficient purification and synthetic approaches. According to present scientific knowledge 2′-O-fucosyllactose is not a crystalline trisaccharide making the development of large scale/low cost manufacturing technologies suitable for the manufacture of high purity 2′-O-fucosyllactose a challenging task.
It is possible to obtain 2′-O-fucosyllactose by four different approaches using isolation, chemical synthesis, enzymatic synthesis and biotechnological methodologies.
The first mentioning of HMO's in the literature appeared in the 1950's. At these times specific human milk oligosaccharides were isolated from human milk by using sophisticated chromatographic protocols. However, the purities of such early isolated samples are rather uncertain due to the high number of human milk oligosaccharide isomers present in mother's milk. For example, 2′-O-fucosyllactose and 3-O-fucosyllactose are both present in human milk and their chromatographic separation have been solved decades later. It has later been found that milk form other mammals also contain some of these oligosaccharides but in extremely low concentrations and ratios different from human milk. In the later years enormous progresses have been made in the production of 2′-O-fucosyllactose via chemical, enzymatic and biotechnological methodologies. However, most of these methodologies have not succeeded in large scale productions providing bulk quantities of 2′-O-fucosyllactose in a commercially attractive price range.
Enzymatic synthesis of 2′-O-fucosyllactose has developed significantly in the last decade by using enzyme cloning/mutating technologies. One specific approach has transferred a fucosidase enzyme into a 1,2-α-L-fucosynthase facilitating the synthesis of glycosidic linkages and avoiding hydrolysis of these at the same time. Unfortunately, 1,2-α-L-fucosynthases are not commercially available in large quantities and therefore to date are not suitable for manufacturing technology developments. A second enzymatic approach has been using α-(1→2)fucosyltransferase, α-(1→2)-L-galactosyltransferase for the creation of interglycosidic linkages. These enzymes are rather sensitive and not really suitable for large scale preparations. Additionally, such enzymes require sugar nucleotide type donors which are hardly available and are extremely expensive. A third enzymatic approach is based upon the use of retaining α-L-fucosidases but the achieved selectivities and yields are usually rather modest.
Either genetically engineered microorganisms or mammals are used in biotechnological methodologies for the synthesis of 2′-O-fucosyllactose. Such technologies use complex enzymatic systems facilitating both the biosynthesis of precursors and the required glycosylations. To date, such approaches face severe regulatory approval hurdles due to the use of genetically engineered organisms and potential contaminations of non-natural oligosaccharides.
Chemical syntheses have until now still been the most economically efficient way to produce 2′-O-fucosyllactose. The hurdles of large scale chemical synthesis are i.e. low stereoselectivities, low overall yields, use of sophisticated and expensive purification methodologies such as column chromatography, and the use of toxic reagents not suitable for food/therapeutic product developments.
In prior art, all syntheses have been using a lactose acceptor and a fully protected L-fucose donor as essential building blocks. The differences among approaches are related to different protecting group strategies, glycosylation methodologies and final purification policies. According to our best knowledge, the first chemical synthesis of 2′-O-fucosyllactose was published in 1981 [1] and since then four further chemical [2-5] and one chemoenzymatic [6] syntheses have been published.
The first chemical synthesis of 2′-O-fucosyllactose was published by K. L. Matta and co-workers in 1981 [1] using a 6-O-benzoylated lactose acceptor and a tri-O-benzylated α-fucopyranosyl bromide donor followed by successive removal of the protecting groups (Scheme 2). The synthesis comprises several chromatographic purification steps in the intermediate stages in order to reach a final intermediate E which contains only benzyl protective groups. Compound E was chromatographed and the purified sample was crystallized from methanol-ether as a dihydrate. However, this dihydrate proved to be rather inconvenient for the last deprotection (catalytic hydrogenolysis) step due to its awkward solubility. Thus the reduction was performed using a dilute solution of compound E in glacial acetic acid. After removing the catalyst the filtrate was evaporated and co-evaporated with methanol, toluene and abs. ethanol several times to afford the crude 2′-FL as a white amorphous solid.

In 1986 M. Martin-Lomas [2] and co-workers published the synthesis of 2′-O-fucosyllactose preparing an isopropylidene-protected final 2′-FL intermediate. The final deprotection step is the cyclic ketal hydrolysis using 20% aqueous acetic acid. The synthetic strategy is utilizing a very rare lactose diol acceptor, which is difficult to make in large quantities. Actually, the precursor itself makes the approach uncompetitive for large scale technology developments. Furthermore, numerous chromatographic purifications are needed for intermediate isolations and the approach doesn't provide crystalline intermediates which might facilitate cheap purification options via crystallization. The final deprotection is a cyclic ketal hydrolysis which is a delicate step due to acid lability of the L-fucose residue. The hydrolytic cleavage condition of the cyclic ketal has to be rather gentle to prevent by-product formations. However, gentle hydrolysis conditions ends up with the presence of unhydrolyzed cyclic ketals in the reaction mixture. Thus, the crude 2′-O-fucosyllactose product requires further chromatographic purifications. Due to the disadvantages listed above the approach is not suitable for manufacturing technology developments.
K. L. Matta and co-workers published additional two syntheses [3, 4] choosing a protection group strategy which facilitated a trifluoroacetic acid assisted hydrolysis of the final 2′-FL intermediate. Both syntheses use the same acceptor but different 4-methoxybenzylated donor molecules. One approach uses n-pentenyl glycoside activated by IDCP while the second approach is based upon thioglycoside activation. The use of pentenyl glycoside with IDCP promoter prevents itself the development of production technologies due to the limited availability of the donor and the high price of the activator. In the case of thioglycoside activation, the fucosylation step provided an α/β-mixture in the ratio (9:1) and a rather difficult chromatographic purification was needed to separate the desired stereoisomer from the unwanted β-product. Thus, none of these methodologies have potentials for multi-ton-scale production of 2′-O-fucosyllactose.
In 1997 H. Hashimoto and co-workers developed a synthesis strategy using acid/oxidation labile protecting groups on both acceptor and donor molecules (isopropylidene and 4-methoxybenzyl) which were removed in trisaccharide intermediate via ceric ammonium nitrate treatment [5]. The approach provides polar products which are rather difficult to handle and contaminated by large quantity of ceric ammonium salts. Both the 2′-O-fucosyllactose and the inorganic impurity were water soluble substances, thus, O-acetylation followed by Zemplén deprotection was needed as the essential purification step of the synthesis. The approach has numerous drawbacks such as use of IDCP as a coupling reagent, removal of p-methoxybenzyl groups using CAN (expensive and toxic) and O-deacetylation of the final 2′-FL intermediate using sodium methoxide in the presence of base sensitive end-product. In general, high purity of 2′-O-fucosyllactose cannot be produced when reducing sugars are treated with strong bases.
In year 2001 L. Lay and co-workers introduced a synthesis of 2′-O-fucosyllactose using both enzymatic and chemical procedures [6]. In this approach a β-benzyllactoside derived acceptor was synthesized using enzymatic manipulations for installing orthogonal protecting groups in a regioselective manner. The last part of the synthetic strategy consists of three steps, the fucosylation, an O-deacetylation and a hydrogenolysis. Unfortunately, the method has never generated commercialization interest due to the relatively modest yields of lipase assisted acylation, the use of toxic and explosive hydrazine derivatives and lack of crystalline intermediates. Multiple chromatographic steps all along the synthesis prevented the development of manufacturing technologies.
In summary, isolation technologies have never been able to provide large quantities of human milk oligosaccharides including 2′-O-fucosyllactose due to the large number of oligosaccharides present in human milk. Additionally, the presence of regioisomers characterized by extremely similar structures further made separation technologies unsuccessful. Enzymatic methodologies suffer from the low availability of enzymes, extremely high sugar nucleotide donor prices and regulatory difficulties due to the use of enzymes produced in genetically modified organisms. The preparation of human milk oligosaccharides via biotechnology has huge regulatory obstacles due to the potential formation of several unnatural glycosylation products. To date, all the chemical methods developed for the synthesis of 2′-O-fucosyllactose have several drawbacks which prevented the preparation of even multigram quantities of 2′-O-fucosyllactose. The most severe drawback of chemical approaches is the lack of design for crystalline intermediates to facilitate low cost purification methodologies and to enhance scale-up opportunities. Thus, the demand for the development of a robust synthetic approach suitable for the production of 2′-O-fucosyllactose has been increasing.
The present invention provides methodologies suitable for both large scale manufacturing and large scale purification of 2′-O-fucosyllactose. The invention is based upon the utilisation of hydrogenolysis of O-benzyl/substituted O-benzyl moieties of novel protected 2′-O-fucosyllactose intermediates. Additionally, it is also an important characteristic of the present invention that the above-mentioned novel O-benzylated/substituted O-benzylated 2′-O-fucosyllactose intermediates have nice crystalline properties assisting powerful purification processes. For example, the realisation of highly crystalline properties of 2′-O-fucosyllactose derivatives allowed the development of powerful manufacturing procedures using entirely crystallisations for product/intermediate purifications. More importantly, the introduction of novel 2′-O-fucosyllactose intermediates provided by the present invention opens the very first opportunities for scale-up options. Before the present invention, complex reaction sequences had to be performed in a continuous manner due to lack of cheap purification options. The novel crystalline 2′-O-fucosyllactose intermediates allow the separation of chemical steps from each other providing real opportunities for scale-up developments. Thus, the crystalline novel intermediates provided by the present invention are responsible for the development of the very first 2′-O-fucosyllactose manufacturing technology suitable to give bulk quantities of high purity 2′-O-fucosyllactose for nutritional and pharmaceutical industries.